Simple value screens seem to work the best. Not only are the returns to simple value screens higher, they’re obviously easier and less time consuming to use as they require fewer inputs. Most importantly, they beat the market!
Wesley Gray, Jack Vogel and Yang Xu tested 13 value screens available on the American Association of Individual Investor’s (AAII) website. They published the results in their paper Does Complexity Imply Value? AAII Value Strategies from 1963 to 2013.
Gray, Vogel and Xu tested the screens using data on all firms listed on the New York Stock Exchange (NYSE), American Stock Exchange (AMEX) and NASDAQ from July 1963 through to December 2013*.
The study focused on large and mid-cap stocks, which the researchers defined as the top 60% percent of market capitalization each year. Excluding small-cap stocks helps to ensure that the test portfolios are liquid. It also helps to ensure that any performance differences aren’t due to the size effect**.
They compared the performance of the 13 value screens to the performance of an equally weighted S&P 500 index. Equal-weighting is the simplest way to introduce a tilt towards value and size, since market capitalization indices assign higher index weights to the stocks of larger and more expensive companies.
Thus an equally-weighted index is a suitable benchmark for a value strategy. It’s also a tougher hurdle to beat as equally-weighted benchmarks usually out-perform market capitalization-weighted benchmarks over the long-term***.
So what did the researchers find? Several of the value screens out-performed, most notably the EBITDA/TEV, Piotroski’s F-score, Greenblatt’s Magic Formula and Price to Free Cash Flow (PFCF).
|Value Screen||Compound Annual Growth Rate||Volatility||Sharpe Ratio|
|S&P 500 EW||12.80%||17.35%||0.497|
The F-Score screen uses 9 criteria, the EBITDA/TEV screen uses 1, the Magic Formula screen uses 2 and the PFCF screen uses 3. With the exception of the F-score, the simplest screens work the best.
There’s a high-correlation between all of the screens, which makes sense since they are all trying to identify cheap stocks, albeit using different screening criteria. The high-correlations between screens also suggest that combining various screens provides little diversification benefit. In other words, one value screen should be enough.
Overall, there isn’t a strategy that outperforms all of the time, but the F-score and EBITDA/TEV screens have the best average performance across sub-samples (1963-1980, 1981-1996 and 1997-2013).
It’s important to remember that screen-based strategies can create highly concentrated portfolios. 8 of the 13 screens tested generate portfolios with a median size below 15 firms. At times the screens may also generate portfolios that are dominated by a particular industry, which will also reduce the diversification benefit of holding several stocks.
Perhaps a better way to use a screen is as an idea generation tool. Let’s assume than an investor would like to create a 30-stock portfolio. One way to do this might be to use a screen every 3 months to select 6 stocks. In other words, our investor progressively adds 6 stocks each quarter until they are fully invested after 18 months.
My grandfather used the same principle when planting his tomatoes. Nonno would plant a row of tomatoes every 2-3 weeks. That way his tomatoes wouldn’t ripen all at once and he’d have fruit for the whole season.
Staggering investment over time has several benefits.
- Having a plan reduces the temptation to “fiddle” with the portfolio, reducing returns due to the effect of transaction costs and taxes.
- Identifying 6 opportunities each quarter is a manageable target.
- By the time the portfolio’s been fully “planted” it might be time to begin “harvesting” some of the ideas that you’ve planted earlier in the season.
- It keeps the portfolio “fresh” as there are always new stocks and ideas coming into the portfolio (but not due to over-trading).
* Excluding REITS, ADRs and closed-end funds.
** There is a large body of academic research demonstrating that small-capitalization stocks usually out-perform large-capitalization stocks over the long-term. Research into the size effect began in 1981 with Rolf Banz paper: The Relationship Between Return and Market Value of Common Stocks and further popularized by the work of Eugene Fama and Ken French in their paper The Cross-Section of Expected Stock Returns.
*** Research by Standard and Poors found that an equally-weighted S&P 500 index out-perform the S&P 500 market capitalization-weighted index over the long-term (although the level of performance also varied considerably under different market conditions).